Portland, Ore. - A four-year skunk works effort at the University of Rhode Island in Kingston has cut the size of an antenna by as much as one-third for any frequency from the kHz to the GHz range. Using conventional components, the four-part antenna design cancels out normal inductive loading, thereby linearizing the energy radiation along its mast and enabling the smaller size.
"The DLM [distributed load monopole] antenna is based on a lot of things that currently exist," said the researcher who invented the smaller antenna, Robert Vincent of the university's physics department, "but I've been able to put a combination of them together to create a revolutionary way of building antennas. It uses basically a helix plus a load coil."
The patent-pending design could transform every antenna-from the GHz models for cell phones to the giant, kHz AM antennas that stud the high ground of metropolitan areas-Vincent said.
For cell phones, for example, Vincent said he has a completely planar design that is less than a third the size of today's cell phone antennas. And those 300-foot tall antennas for the 900-kHz AM band that dominate skylines would have to be only 80 feet high, with no compromise in performance, using Vincent's design, he said.
"When looking at these antennas, you pretty much have to forget everything you ever knew about antennas and keep an open mind, because some of the things I have done are very radical," said Vincent. "With my technique, I reduce the inductive loading that is normally required to resonate the antenna by as much as 75 percent . . . by utilizing the distributed capacitance around the antenna."
Vincent, an amateur radio operator, embarked on his project after he moved to a new neighborhood and his neighbors objected to the 140-foot tall antenna he planned to erect for a quarter-wave 1.8-MHz transmitter. So he surveyed the literature, took the best of the best designs and combined them into a 21-MHz test antenna that was 18 inches high, as opposed to the 12- to 24-foot height of the antennas normally used for that band. Building on that work, he eventually devised a 46-foot-tall 1.8-MHz antenna his neighbors could accept.
"I looked at all the different approaches used to make antennas smaller, and there seemed to be good and bad aspects" to each, Vincent said. "A helix antenna is normally known to be a core radiator, because the current profile drops off rapidly; they are just an inductor, and inductance does not like to see changes in current, so it's going to buck that. "But what I found was that for any smaller antenna, if you place a load coil in the middle you can normalize and make the current through the helix unity; that is, you can maximize it and linearize it."
Vincent has verified designs from 1.8 MHz to 200 MHz by measuring and characterizing the behavior of his DLM antenna compared with a normal quarter-wave antenna of the same frequency. He found that many of the disadvantages of traditional antennas were not problems for the much lighter inductive loading in a DLM.
"For instance, in a normal quarter-wave antenna the current continually drops off in a sinusoidal shape, but these antennas don't do that," said Vincent. "The current at the top of the antenna is 80 percent of the current at the base."
The reason more current can be pumped into a DLM design than in a conventional equivalent at the same size, Vincent theorized, is that the DLM distributes energy more evenly along the antenna's length. Using a DLM antenna one-third to one-ninth the size of standard quarter-wave antenna, he measured nearly 80 percent efficiency, when conventional wisdom would dictate that an antenna the size of a DLM should be only 8 to 15 percent efficient.
To check his theory, Vincent analyzed and compared the current profiles, output power and a score of other standard tests for measuring antenna performance. All measurements were in reference to comparative measurements made on a quarter-wave vertical antenna for the same frequency, on the same ground system and same power input.
"I was able to increase the current profile of the antenna over a quarter-wave by as much as two to 2.5 times," said Vincent. "That is, the magnitude of the current in these antennas is two to 2.5 times larger than for a normal quarter-wave antenna.
"However, if you measure the current profiles for both antennas and integrate the area under the curves, you come out with the same volume, indicating that the much smaller antenna is filling the airwaves with the same amount of radio energy."
Vincent plans to publish the results in a scientific journal soon, but with a patent decision imminent, he couldn't hold off a preliminary announcement that his theories regarding DLM antennas were being supported by the experimental results. According to the researcher, the DLM antenna profiles look just like the theoretically ideal antenna profile-operating on a single frequency with very high efficiency, while not producing any interfering frequencies or wasting thermal energy.
"The phase and amplitude of this antenna are a perfect mimic of the universal resonance curve," said Vincent. "This makes the antenna completely predictable well beyond its bandwidth. Another unique feature is that these antennas have no harmonic response whatsoever; as a matter of fact, to a certain extent I used filter synthesis to design the antennas."
To the naked eye, the DLM antenna looks unremarkable, said Vincent, who jokes that you could put a flag on his antennas and they would look like flagpoles. But under the skin are four main sections to the antenna (from bottom to top): an inductive helix, a capacitive midsection, an inductive load coil and a capacitive top section. The different lengths of the mid- and top sections give them different resonant frequencies, which, together with the exact values of inductance and capacitance, define the antennas design specifications for any desired frequency.
"The technology is completely scalable: Take the component values and divide them by two, and you get twice the frequency; take all the component values and multiply them by two, and you are at half the frequency," said Vincent. "There are two poles in the antenna, and where I place the poles in relation to one another-how much I bring the two resonant frequencies together or spread them apart-enables me to emulate different antennas, from a quarter-wave to a five-eighths wave."
Vincent said no existing modeling software could adequately model his antenna design. So he rolled his own simulation with Mathcad, making use of some of Mathcad's filter design algorithms for the inductive/capacitive-canceling effect.
"Eight years ago, antenna design was 90 percent black magic and 10 percent theory," said Vincent. "But now, with my design, they are 10 percent black magic and 90 percent theory."
The antennas are also well-behaved, with wide bandwidth and easy to connect to standard equipment, according to Vincent. For instance, they can directly connect to standard 50-ohm antenna inputs without any adapters.
"All I have to do is tap the helix at its base, and you get a perfect 50-ohm match with out any lossy networks [as are required for other advanced antenna designs]," said Vincent.
For the future, Vincent is moving up into the GHz bands for use with cell phones and radio-frequency ID equipment. A problem in the past has been that as components are downsized, they become too small to utilize standard antenna materials. At 1 GHz, for example, the helix is only eight-thousandths of an inch in diameter and requires more than 100 turns of wire.
"So I came up with a new way of developing a helix for high frequencies that is a fully planar design; it's a two-dimensional helix," said Vincent.
With the new helix design, Vincent has built a prototype 7-GHz antenna that he claims is indistinguishable from a quarter-wave antenna in all but its size. "Because the new design is completely planar, we could crank these out using thin-film technologies," Vincent said.
Vincent received the 2004 Outstanding Intellectual Property Award from the University of Rhode Island's Research Office, joint applicant for the patent.